Differentiable Causal Backdoor Discovery

Limor Gultchin, Matt Kusner, Varun Kanade, Ricardo Silva
Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, PMLR 108:3970-3979, 2020.

Abstract

Discovering the causal effect of a decision is critical to nearly all forms of decision-making. In particular, it is a key quantity in drug development, in crafting government policy, and when implementing a real-world machine learning system. Given only observational data, confounders often obscure the true causal effect. Luckily, in some cases, it is possible to recover the causal effect by using certain observed variables to adjust for the effects of confounders. However, without access to the true causal model, finding this adjustment requires brute-force search. In this work, we present an algorithm that exploits auxiliary variables, similar to instruments, in order to find an appropriate adjustment by a gradient-based optimization method. We demonstrate that it outperforms practical alternatives in estimating the true causal effect, without knowledge of the full causal graph.

Cite this Paper


BibTeX
@InProceedings{pmlr-v108-gultchin20a, title = {Differentiable Causal Backdoor Discovery}, author = {Gultchin, Limor and Kusner, Matt and Kanade, Varun and Silva, Ricardo}, booktitle = {Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics}, pages = {3970--3979}, year = {2020}, editor = {Chiappa, Silvia and Calandra, Roberto}, volume = {108}, series = {Proceedings of Machine Learning Research}, month = {26--28 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v108/gultchin20a/gultchin20a.pdf}, url = {https://proceedings.mlr.press/v108/gultchin20a.html}, abstract = {Discovering the causal effect of a decision is critical to nearly all forms of decision-making. In particular, it is a key quantity in drug development, in crafting government policy, and when implementing a real-world machine learning system. Given only observational data, confounders often obscure the true causal effect. Luckily, in some cases, it is possible to recover the causal effect by using certain observed variables to adjust for the effects of confounders. However, without access to the true causal model, finding this adjustment requires brute-force search. In this work, we present an algorithm that exploits auxiliary variables, similar to instruments, in order to find an appropriate adjustment by a gradient-based optimization method. We demonstrate that it outperforms practical alternatives in estimating the true causal effect, without knowledge of the full causal graph.} }
Endnote
%0 Conference Paper %T Differentiable Causal Backdoor Discovery %A Limor Gultchin %A Matt Kusner %A Varun Kanade %A Ricardo Silva %B Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2020 %E Silvia Chiappa %E Roberto Calandra %F pmlr-v108-gultchin20a %I PMLR %P 3970--3979 %U https://proceedings.mlr.press/v108/gultchin20a.html %V 108 %X Discovering the causal effect of a decision is critical to nearly all forms of decision-making. In particular, it is a key quantity in drug development, in crafting government policy, and when implementing a real-world machine learning system. Given only observational data, confounders often obscure the true causal effect. Luckily, in some cases, it is possible to recover the causal effect by using certain observed variables to adjust for the effects of confounders. However, without access to the true causal model, finding this adjustment requires brute-force search. In this work, we present an algorithm that exploits auxiliary variables, similar to instruments, in order to find an appropriate adjustment by a gradient-based optimization method. We demonstrate that it outperforms practical alternatives in estimating the true causal effect, without knowledge of the full causal graph.
APA
Gultchin, L., Kusner, M., Kanade, V. & Silva, R.. (2020). Differentiable Causal Backdoor Discovery. Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 108:3970-3979 Available from https://proceedings.mlr.press/v108/gultchin20a.html.

Related Material